A buddy of mine recently tweeted something about rodent damage to his car, and it being expensive to fix, and that reminded me to get off my duff and put up this post recounting my brush with automotive mouse damage earlier this year. Mice like confined, dark, warm areas and cars are just full of those. Some of the worst stories involve mice chewing through air filters in the airbox, and the debris then getting sucked directly into the engine and causing internal engine damage. Obviously expensive, but then also a very obvious problem. At the other end of the critter damage spectrum is wiring damage. One one hand, a damaged wire or connector can be easily fixed (ok, so perhaps your mechanic would ding you with replacing an entire wiring harness, which could be expensive, but technically you can splice to repair the damage for very little cost), but on the other hand the symptoms can literally be anything and the root cause of the problem can be extremely time consuming to find, often requiring special equipment. If your mechanic spends all day chasing down electrical gremlins, that is already quite a repair bill just in labor. And often, since many shops are under such pressure to fix a car and get it out the door, the job will involve swapping out a lot of parts that are actually fine, further adding to the repair bill (the extra labor and, if the shop is trying to pull a fast one, they may even charge you for new parts that weren't necessary). Let us take my latest mousecapade as an example, and along the way we shall consider how things might have gone down, had I taken the car to a shop.

This story involves my Mk5 R32. Our neighbors across the street, wonderful folks, have a son who is a real gearhead and worked as a mechanic for a while. He appreciates cars, and I know he loves R32s. He was stationed out west and when I heard he was coming home for a visit I thought to myself "Wow he hasn't seen the R32 yet... I bet he'd like to take it for a drive". So he and I got to chatting outside one day and I told him he was welcome to take her for a spin if he'd like... of course, he took me up on the offer. We both got in, I gave him the rundown on how the DSG works, how to switch off the traction and stability control, etc. He went to start the car... and nothing. Dash lights on, but no attempt made by the starter to turn over the engine. I said "buckle your seatbelt, this car might have one of those safety interlocks" (I was pretty sure it didn't, but this was my hopeful thinking). He buckled, tried again, and the engine roared to life... then died after a split second. Now the dash was lit up like a Christmas tree and again the engine wouldn't turn over. This guy felt just awful, because he knew all too well how I must have been feeling at that moment, and even though he must have known deep down it was luck of the draw rather than anything that was his fault, he still felt like he had broken my baby.

During some previous work I had noticed a mouse nest underneath the raintray cover, that I had removed... so I immediately started to suspect some kind of rodent damage. Keep that in mind, that if that hadn't already been in my mind this whole thing could have drawn out even longer. This is the first CAN bus car I've owned, so I had to get an updated version of Ross-Tech's interface to be able to read all of the controller modules. While I was waiting for the new interface to arrive, I spent a bunch of time physically inspecting all the wiring I could get to. I tore down the rain tray and engine bay enough to inspect all the wiring and connecotrs there, I removed the bellypan to inspect everything unerneath the car, I stripped everything out of the dash to inspect as much as I could of the wiring in the dash and footwells, and I stripped out the rear luggage compartment to inspect everything in there. Absolutely no sign of any infestation, everything was squeaky clean. It took several days of free time to do all that. When my new vag-com interface arrived, I hooked it up and did a scan. There was an unmanageable number of faults logged, across all of the different modules, with no apparent rhyme or reason to them. I saved them, cleared the system, then attempted to start the car again. The engine fired up ran well. The check engine light was on and the traction control and all wheel drive were disabled. I performed another scan and this time got a much shorter list of faults, most of them complaining about the controller for the Haldex AWD module being offline. The Haldex unit is about $800, if you catch a deal on one... I certainly was not looking forward to replacing it. But with the wiring having been inspected, I was starting to come to the conclusion that it must be the Haldex unit. They are notorious for going bad, especially in areas like New England, where the climate accelerates corrosion around the connector, eventually working its' way inside the module itself. I inspected the connector on my module, and it was beautifully clean. Let's pause for a moment and consider what a shop might have done if they were working on the car. They would have started with a system scan, of course, although the car would have needed towing to the shop because it was immobile to start with. Right away you've paid for towing, and for diagnostics scan. They would have been faced with the wall of fault codes and then would have either cleared them and re-scanned (as I did), or just gone ahead and assumed that the problem was the Gateway controller module (the common link between all the CAN bus modules), and either replaced it or spent a lot of time troubleshooting it. They might have even replaced it, re-scanned, found that most of the codes were gone, and thought that they had legitimately fixed part of the problem. We're well over $1500 here so far (the gateway module is very expensive, difficult to get to, and requires special equipment to code to the rest of the system so that it will function properly), and nothing has actually been fixed or diagnosed yet. Now, faced with codes mostly pointing to the Haldex controller being offline, they might have been diligent enough to at least check for a good ground at the Haldex module plug, as bad grounds are also a common problem. In my case, the ground checked out fine. To test the CAN bus signal wires would take something better than a common multimeter, I sure didn't have the means to test them. At this point a shop could very well justify replacing the Haldex module, which is a bit of a pain if you don't either have some special or modified tools, or drop down the rear drivetrain a little bit. Even once they found that the replacement module didn't make any difference, they could easily avoid the hassle and cost of swapping the parts back out by simply charging the customer for a new module anyway. Most customers wouldn't know any better than to throw up their hands and curse Volkswagen's ineptitude over the supposed failure of such an expensive part. Two things were holding me back from ordering a new Haldex module. First, of course, was the price. Second, was the fact that there were some other seemingly unrelated fault codes that I'm sure popped up at exactly the same time. And I wasn't about to spend $800+ on a part when coincidence was part of the equation. I had other symptoms and fault codes logged that turned out to be caused by my own foolishness... I had placed one of the big power distribution fuses in the engine bay back in the wrong spot when I was doing that first round of inspection and troubleshooting, and that was causing my fresh air blower, sunroof, and headlight washers to stop working (and throw associated codes, of course). Even with that addressed, I had a strange beahavior with my fuel guage. It would only read up to, but not beyond, halfway. With a full tank of gas, the gauge would stay at half full until the actual level dropped below half, and then it would read the correct amount. There were no faults stored that seemed related to the fuel sender. How could that be related to the Haldex module going down? Electrically, they had nothing directly in common. It was about this time that someone replied to a thread I had started over on the Ross-Tech forums, suggesting that I take a look under my back seat. This fellow said that the R32 has two fuel pumps and two fuel level sending units. On one side of the car, the wiring for that fuel pump and sending unit physically runs right next to the wiring going to the Haldex control module. That sounded fantastic to me, so I took up the rear seat and removed the cover plate for the pump on the passenger side... everything looked perfect. Under the cover plate on the driver side, though, was an entirely different story. Here is what I found:

Some little furry friends had made a nice, cozy home in there complete with all the terrible tasting licorice they could eat. Both wiring harnesses had been chewed through, there were only a couple of wires hanging on by a strand or two of copper. The little rascals didn't even have the courtesy to leave me enough wiring out of the connectors to splice back in. They had chewed most of the wires right up into the back sides of the connectors, so that I was faced with un-pinning each connector, uncrimping each pin, then re-crimping it onto a freshly stripped wire. A few of the wires I was able to solder, barely. Here again, once a repair shop had found this damage, they would not have wasted their time with a repair, they would have ordered whole new wiring harnesses. It looks like to replace those harnesses you'd have to drop down the fuel tank and the whole rear end a little bit, so I'm sure it would be a few hours of labor on top of the already steep cost of the harnesses themselves. Had a shop started with that seemingly unrelated fuel gauge quirk as the basis of their investigation, then they would have quickly found the real source of the problem. But worst case a customer could have been easily gouged for both replacement harnesses, replacement CAN bus gateway module, and replacement Haldex module (the new ones do not come with the Haldex clutch solenoid, so a new seal kit for that would be necessary also) as well as a couple of days labor (if they had burned most of a day chasing down the wiring and getting nowhere, as I did). This could have been a $4000 repair bill. It took a lot of blood, sweat, and tears, but I got through it without paying any cash and everything has been working perfect since then. The exposed copper wiring, where it had been chewed through, was corroded enough that I think the mouse (or mice) had been there for quite some time.

In summary, what appears to have happened was when the last of the wiring for the Haldex module finally let go (or was chewed through), it must have spiked the CAN bus signal lines and caused all the other modules to momentarily lose their grip on reality. I'm pretty lucky none of the modules were damaged, I think. For the morbidly curious, here is my scan after clearing the initial set of codes from "the incident". This was the set of relevant codes: